Deep learning-based smart speaker to confirm surgical sites for cataract surgeries: A pilot study

Version 1.01
In the experiments, we attempted to add additional short words from the various text-to-voice tools. The researchers recorded the target words, such as time-out, cataract, phacoemulsification, and intraocular lens, with varying accents, speed, and voice tones that provided by the text-to-voice tools. As the voice interface relied on keyword spotting to initialize the interactions in most devices, “time-out” was assigned as a keyword to initialize the automated detection. Finally, the dataset consists of different people speaking the same word for training and validation. The Speech Commands dataset provides several basic noise data including background sounds from white noise, pink noise, exercise, and doing the dishes. Additional noise sounds in operation room including vital monitoring sound and background sound of surgery were added in the noise database.

版本1.01
本实验中，我们尝试从各类文本转语音（Text-to-Voice）工具中采集额外的短句词汇。研究人员录制了由各类文本转语音工具生成的目标词汇，涵盖time-out、白内障（cataract）、超声乳化术（phacoemulsification）与人工晶状体（intraocular lens），且这些语音样本带有不同的口音、语速与语调。由于多数智能设备的语音界面均依赖关键词检测（Keyword Spotting）来启动交互，因此将“time-out”设为启动自动检测的触发关键词。最终，本数据集包含不同说话人录制的同一目标词汇，用于模型训练与验证。语音命令数据集（Speech Commands Dataset）提供了多类基础噪声数据，涵盖白噪声、粉噪声、运动场景及洗碗场景的背景音效。此外，我们还向噪声库中补充了手术室场景的相关噪声，包括生命体征监测音及手术背景音。